A diagnosis method, device and equipment of a search service and a storage medium

Abstract

Embodiments of the present application disclose a search service diagnosis method and device, equipment and a storage medium, related embodiments can be applied to cloud technology, artificial intelligence, intelligent transportation and various other scenarios, for improving the diagnosis efficiency of the search service. The method of the embodiments of the present application comprises: obtaining a target search statement for a target search service, recalling a first search result of each search resource from a plurality of search resources according to the target search statement, wherein the plurality of search resources belong to search resources associated with the target search service, sorting the first search result of each search resource to obtain a second search result of each search resource, diagnosing the processing condition of each search resource in each stage in the plurality of search resources according to the second search result of each search resource to obtain diagnosis information of the target search service, wherein the diagnosis information comprises first diagnosis information for the recall stage and second diagnosis information for the sorting stage.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of artificial intelligence technology, and in particular to a search service diagnosis method and device, equipment and a storage medium. BACKGROUND

[0002] With the continuous development of Internet technology and information processing technology, more and more terminal users access the Internet, and product service developers have developed various Internet products to help more and more terminal users search for text, emoticons or image information in daily work, study and life using various Internet products. In order to make the Internet product better serve the terminal user, the research and development personnel obtain the diagnosis result by diagnosing and analyzing the process of the terminal user searching the Internet product to more accurately locate the problems existing in the Internet product, and provide technical operation guarantee and improvement basis for the Internet product.

[0003] Currently, the diagnosis and analysis of the search process of the Internet product is usually performed by searching for the corresponding search log in the product log, extracting relevant feature information from the search log, and then performing artificial diagnosis and analysis on the relevant feature information to obtain the diagnosis result.

[0004] However, the extraction of relevant feature information from the search log is easily affected by factors such as information missing, information singleness and extraction error, resulting in obtaining inaccurate or incomplete relevant feature information of the search process. Moreover, diagnosis and analysis based on relevant feature information requires a large amount of human cost to analyze and process massive data, resulting in increased time cost, thereby leading to low accuracy and efficiency of the obtained diagnosis result. SUMMARY

[0005] Embodiments of the present application provide a search service diagnosis method, device, equipment and storage medium for recalling the first search result of each search resource at one time, which can avoid the influence of factors such as information missing and extraction error in the search log, and does not need to repeatedly search the log for feature mining and diagnosis for multiple search resources. The first diagnosis information for the recall stage and the second diagnosis information for the sorting stage can be directly diagnosed based on the second search result, thereby improving the accuracy and efficiency of the search result and the diagnosis information.

[0006] In one aspect, the embodiments of the present application provide a search service diagnosis method, comprising:

[0007] obtaining a target search statement for a target search service;

[0008] recalling a first search result of each search resource from a plurality of search resources according to the target search statement, wherein the plurality of search resources belong to search resources associated with the target search service;

[0009] rank the first search result of each search resource to obtain a second search result of each search resource;

[0010] diagnose, according to the second search result of each search resource, a processing condition of each search resource in each stage to obtain diagnosis information of the target search service, wherein the diagnosis information comprises first diagnosis information for the recall stage and second diagnosis information for the ranking stage.

[0011] Another aspect of the present application provides a diagnosis apparatus of a search service, comprising:

[0012] a obtaining unit configured to obtain a target search sentence for a target search service;

[0013] a processing unit configured to recall, according to the target search sentence, a first search result of each search resource from a plurality of search resources, wherein the plurality of search resources belong to search resources associated with the target search service;

[0014] the processing unit is further configured to rank the first search result of each search resource to obtain a second search result of each search resource;

[0015] the processing unit is further configured to diagnose, according to the second search result of each search resource, a processing condition of each search resource in each stage to obtain diagnosis information of the target search service, wherein the diagnosis information comprises first diagnosis information for the recall stage and second diagnosis information for the ranking stage.

[0016] In a possible design, in an implementation manner of the another aspect of the embodiment of the present application, the processing unit can be specifically configured to:

[0017] perform word segmentation processing on the target search sentence to obtain at least two to-be-processed words;

[0018] perform semantic recognition operation on the at least two to-be-processed words respectively to obtain a semantic recognition probability corresponding to each to-be-processed word;

[0019] determine an industry category of the target search sentence according to the semantic recognition probability corresponding to each to-be-processed word;

[0020] obtain a first search result of the target search sentence in each search resource according to the industry category of the target search sentence.

[0021] In a possible design, in an implementation manner of the another aspect of the embodiment of the present application, the processing unit can be specifically configured to:

[0022] read a resource database to obtain a scoring strategy corresponding to each search resource;

[0023] score each search result in the first search result of each search resource according to the scoring strategy, to obtain a recall score of each search result in the first search result of each search resource;

[0024] sort the first search result of each search resource according to the recall score of each search result, to obtain a second search result of each search resource.

[0025] In a possible design, in an implementation of the other aspect of the embodiment of the present application,

[0026] The obtaining unit is further configured to obtain a synchronous reporting path according to a synchronous reporting parameter, where the synchronous reporting parameter is used to control the synchronous reporting switch to open the synchronous reporting path.

[0027] The processing unit is further configured to upload the first search result and the second search result of each search resource and the diagnostic information of the target search service to the diagnostic database through the synchronous reporting path.

[0028] In a possible design, in an implementation of the other aspect of the embodiment of the present application,

[0029] The processing unit is further configured to perform standardization processing on the first search result and the second search result of each search resource and the diagnostic information of the target search service respectively, to obtain a first standard search result and a second standard search result of each search resource and standard diagnostic information of the target search service.

[0030] The processing unit can be specifically configured to upload the first standard search result and the second standard search result of each search resource and the standard diagnostic information of the target search service to the diagnostic database through the synchronous reporting path.

[0031] In a possible design, in an implementation of the other aspect of the embodiment of the present application,

[0032] write the first search result and the second search result of each search resource and the diagnostic information of the target search service into a target request packet to obtain a target request packet;

[0033] upload the target request packet to the diagnostic database through the synchronous reporting path.

[0034] In a possible design, in an implementation of the other aspect of the embodiment of the present application,

[0035] The processing unit is further configured to perform syntax tree analysis on the first search result and the second search result of each search resource and the diagnostic information of the target search service, to obtain a first analysis result and a second analysis result of each search resource and diagnostic analysis information of the target search service.

[0036] The generating unit is configured to generate a search diagnosis list of the target search service according to the first analysis result, the second analysis result of each search resource, and diagnosis analysis information of the target search service according to a list generation template.

[0037] In a possible design, in an implementation of another aspect of the embodiment of the present application,

[0038] The obtaining unit is further configured to obtain an asynchronous reporting path according to an asynchronous reporting parameter, where the asynchronous reporting parameter is used to control the asynchronous reporting switch to open the asynchronous reporting path.

[0039] The processing unit is further configured to upload the first search result, the second search result of each search resource, and the diagnosis information of the target search service to the diagnosis database through the asynchronous reporting path.

[0040] In a possible design, in an implementation of another aspect of the embodiment of the present application,

[0041] The processing unit can specifically be configured to:

[0042] aggregate the first search result, the second search result of each search resource, and the diagnosis information of the target search service to obtain a diagnosis data set of the target search service;

[0043] write the diagnosis data set of the target search service to the first message queue through the asynchronous reporting path;

[0044] perform aggregation processing on the diagnosis data set of the target search service in the first message queue according to a time dimension to obtain aggregation diagnosis data corresponding to each time window of the target search service;

[0045] In a possible design, in an implementation of another aspect of the embodiment of the present application,

[0046] The obtaining unit is further configured to obtain a current resource consumption corresponding to the target search service;

[0047] The processing unit is further configured to perform data compression on the diagnosis data set of the target search service to obtain diagnosis compressed data of the target search service if the current resource consumption corresponding to the target search service is greater than or equal to the resource threshold.

[0048] The processing unit can specifically be configured to: write the diagnosis compressed data of the target search service to the first message queue through the asynchronous reporting path.

[0049] In a possible design, in an implementation of another aspect of the embodiment of the present application, the processing unit can be specifically configured to:

[0050] obtain target routing information of the target search service according to the search identifier, where the target routing information is obtained by hashing the search identifier;

[0051] route the first search result, the second search result and the diagnosis information of the target search service of each search resource to the target node corresponding to the target routing information for aggregation, to obtain a diagnosis data set of the target search service.

[0052] In a possible design, in an implementation of another aspect of the embodiment of the present application,

[0053] The processing unit is further configured to index the delay reporting data and the target node change data according to the search identifier.

[0054] The processing unit is further configured to aggregate the delay reporting data and the target node change data, to obtain a delay data set.

[0055] The processing unit can be specifically configured to: write the diagnosis data set and the delay data set of the target search service to the first message queue through an asynchronous reporting path.

[0056] Another aspect of the present application provides a computer device, which comprises a memory, a transceiver, a processor and a bus system;

[0057] The memory is configured to store a program.

[0058] The processor is configured to implement the method of each aspect described above when executing the program in the memory.

[0059] The bus system is configured to connect the memory and the processor, so that the memory and the processor communicate.

[0060] Another aspect of the present application provides a computer readable storage medium, which stores instructions, and when the instructions run on a computer, the computer executes the method of each aspect described above.

[0061] As can be seen from the above technical solutions, the embodiment of the present application has the following advantages:

[0062] By acquiring a target search statement for a target search service, a first search result of each search resource is recalled from a plurality of search resources associated with the target search service according to the target search statement, the first search result of each search resource can be sorted to obtain a second search result of each search resource, and then the processing of each search resource in each stage can be diagnosed according to the second search result of each search resource to obtain the diagnosis information of the target search service including the first diagnosis information for the recall stage and the second diagnosis information for the sorting stage. Through the above-mentioned manner, the first search result of each search resource can be recalled from a plurality of search resources associated with the target search service directly and at one time according to the target search statement, which can avoid the influence of information missing, extraction error and other factors in the search log, and does not need to repeatedly search the log for feature mining and diagnosis for a plurality of search resources. The second search result can be obtained based on the first search result of each search resource recalled at one time, and the first diagnosis information for the recall stage and the second diagnosis information for the sorting stage can be directly diagnosed based on the second search result, thereby improving the accuracy and efficiency of the search result and the diagnosis information. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 is an architecture schematic diagram of a business object control system in an embodiment of the present application;

[0064] Figure 2 is an embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0065] Figure 3 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0066] Figure 4 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0067] Figure 5 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0068] Figure 6 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0069] Figure 7 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0070] Figure 8 is another embodiment flowchart of a diagnosis method of a search service in an embodiment of the present application;

[0071] Figure 9is another embodiment flow chart of the diagnosis method of the search service in the embodiments of the present application;

[0072] Figure 10 is another embodiment flow chart of the diagnosis method of the search service in the embodiments of the present application;

[0073] Figure 11 is another embodiment flow chart of the diagnosis method of the search service in the embodiments of the present application;

[0074] Figure 12 is a principle flow schematic diagram of the diagnosis method of the search service in the embodiments of the present application;

[0075] Figure 13 is an embodiment schematic diagram of the diagnosis apparatus of the search service in the embodiments of the present application;

[0076] Figure 14 is an embodiment schematic diagram of the computer device in the embodiments of the present application. DETAILED DESCRIPTION

[0077] The embodiments of the present application provide a diagnosis method, apparatus and device of a search service and a storage medium, which are used for recalling first search results of each search resource at one time, can avoid the influence of information missing, extraction error and other factors in the search log, do not need to repeatedly search the log for feature mining and diagnosis for multiple search resources, can directly diagnose the first diagnosis information for the recall stage and the second diagnosis information for the sorting stage based on the second search results, and thereby improve the accuracy and efficiency of the search results and the diagnosis information.

[0078] The terms "first", "second", "third", "fourth" and the like (if any) in the specification and claims of the present application and the drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a particular order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented, for example, in an order other than that illustrated or described herein. In addition, the terms "include" and "correspond to" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0079] With the rapid development of information, artificial intelligence (AI) gradually enters all aspects of people's life. Artificial intelligence has wide practical significance in text translation, intelligent question answering, sentiment analysis and other aspects. The emergence of artificial intelligence has greatly facilitated people's life. First, a brief description of artificial intelligence. Artificial intelligence is to use mathematical computer or digital computer controlled machine to simulate, extend and expand human intelligence, perceive environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which tries to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machine has the functions of perception, reasoning and decision-making.

[0080] Artificial intelligence is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. The basic technology of artificial intelligence generally includes sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system and mechatronics technology. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning technology. In the machine learning / deep learning technology, cloud technology is often used.

[0081] Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, application technology and other technologies based on cloud computing business model application, which can form a resource pool, be used on demand and be flexible and convenient. Cloud computing technology will become an important support. The background service of the technical network system needs a large amount of computing and storage resources, such as video websites, picture websites and more portals. With the high development and application of Internet industry, every item may have its own identification mark in the future, which needs to be transmitted to the background system for logical processing. Different levels of data will be processed separately, and various industry data need strong system support, which can only be realized through cloud computing.

[0082] Among them, cloud security (Cloud Security) refers to the general term of security software, hardware, users, institutions, security cloud platform based on cloud computing business model application. Cloud security integrates emerging technologies and concepts such as parallel processing, grid computing, unknown virus behavior judgment, etc. Through the abnormal monitoring of software behavior in the network by a large number of clients in the network, the latest information of Internet viruses and malicious programs is obtained, and sent to the server for automatic analysis and processing, and then the virus and Trojan solution is distributed to each client. The search business diagnosis method provided by the embodiment of the application can be realized through cloud computing technology and cloud security technology.

[0083] It should be understood that the diagnostic method for search services provided in this application can be applied to fields such as cloud technology, artificial intelligence, and intelligent transportation, and can be used to locate problems in internet products through diagnostic information. For example, it can locate problems with images searched by search engine A by obtaining diagnostic information. Another example is locating problems with articles searched by search mini-program B by obtaining diagnostic information. Yet another example is locating problems with videos searched by search software C by obtaining diagnostic information. In these scenarios, to obtain diagnostic information, the usual method is to search for relevant search logs in the product logs, extract relevant feature information from the search logs, and perform manual diagnostic analysis to obtain diagnostic results. However, extracting relevant feature information from search logs is easily affected by factors such as missing information, limited information, and extraction errors, leading to inaccurate or incomplete relevant feature information from the search process, resulting in low accuracy and efficiency of the obtained diagnostic results.

[0084] To address the aforementioned issues, this application proposes a diagnostic method for search services, which is applied to... Figure 1 Please refer to the business object control system shown. Figure 1 , Figure 1 This is a schematic diagram of the architecture of the business object control system in an embodiment of this application, such as... Figure 1 As shown, the server obtains the target search statement for the target search business, and recalls the first search result of each search resource from multiple search resources associated with the target search business based on the target search statement. The first search results of each search resource can be sorted to obtain the second search result of each search resource. Then, based on the second search result of each search resource, the processing status of each search resource at each stage can be diagnosed, obtaining diagnostic information for the target search business, including first diagnostic information for the recall stage and second diagnostic information for the sorting stage. This method allows for the direct, one-time recall of the first search result of each search resource from multiple search resources associated with the target search business based on the target search statement. This avoids the impact of missing information or extraction errors in the search logs, eliminating the need for repeated log searches for feature mining and diagnosis of multiple search resources. The second search result can be obtained based on the one-time recall of the first search result of each search resource, and the first diagnostic information for the recall stage and the second diagnostic information for the sorting stage can be directly diagnosed based on the second search result, thereby improving the accuracy and efficiency of search results and diagnostic information.

[0085] Understandable Figure 1Only one terminal device is shown in the figure, but in actual scenarios, more types of terminal devices can participate in the data processing process, including but not limited to mobile phones, computers, smart voice interaction devices, smart home appliances, vehicle-mounted terminals, and the like, and the specific number and types depend on the actual scenario, which is not limited here. In addition, Figure 1 A server is shown in the figure, but in actual scenarios, multiple servers can also participate, especially in multi-model training interaction scenarios, and the number of servers depends on the actual scenario, which is not limited here.

[0086] It should be noted that in the present embodiment, the server can be a standalone physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms, and the like. Basic cloud computing services. The terminal device and the server can be directly or indirectly connected through wired or wireless communication, and the terminal device and the server can be connected to form a blockchain network, which is not limited in the present application.

[0087] To solve the above problems, the present application provides a search service diagnosis method, which is generally executed by a server or a terminal device. Correspondingly, the search service diagnosis apparatus is generally provided in the server or the terminal device.

[0088] It can be understood that the search service diagnosis method, apparatus, device and storage medium disclosed in the present application, wherein the plurality of servers or terminal devices can be composed of a blockchain, and the server or terminal device is a node on the blockchain. In actual application, data sharing between nodes can be required in the blockchain, and each node can store business diagnosis data, search data, etc.

[0089] The search service diagnosis method of the present application will be introduced below. Please refer to Figure 2 The search service diagnosis method of the present application includes one embodiment:

[0090] In step S101, a target search statement for a target search service is obtained;

[0091] In this embodiment, when it is necessary to verify the service capability of the target search service or locate problems existing in the process in which the target search service provides the search service, the target search statement for the target search service can be acquired, so that the search result and the diagnostic information of each search stage corresponding to the search result can be acquired based on the target search statement subsequently, thereby improving the efficiency of acquiring the diagnostic information to a certain extent.

[0092] The target search service can be a search website, a search engine, search software, a search applet, or other search services, and is not limited here.

[0093] Specifically, as shown in Figure 12 When it is necessary to verify the service capability of the target search service or locate problems existing in the process in which the target search service provides the search service, the developer of the target search service can perform a diagnostic selection operation of the target search service on the terminal device, the terminal device can perceive and generate a search diagnostic instruction corresponding to the diagnostic selection operation, so that the server can receive the search diagnostic instruction and acquire the target search statement selected or input by the developer through the terminal device according to the search diagnostic instruction; or the search diagnostic instruction carries a search identifier, so that the server can acquire the target search statement corresponding to the search identifier from the database according to the search identifier, wherein the search identifier is a statement identifier code used to indicate the target search statement, and can be an integer (int) type of digital string or a string; or the search diagnostic client is installed on the terminal device, and the search target object account, the geographic location, the experimental parameter, the access background environment, the Query and other parameters are set on the client, wherein the search target object account is used to indicate the identity account of the developer, and can maintain the security of the search diagnostic process through identity security verification, so that the developer of the target search service can perform the diagnostic selection operation of the target search service on the client, the terminal device can perceive and generate a search diagnostic instruction corresponding to the diagnostic selection operation, so that the search background of the client can acquire the target search statement selected or input by the developer through the client according to the search diagnostic instruction, and the target search statement of the target search service can also be acquired through other manners, which is not limited here.

[0094] In step S102, the first search result of each search resource is recalled from a plurality of search resources according to the target search statement, wherein the plurality of search resources belong to the search resources associated with the target search service;

[0095] In this embodiment, after obtaining the target search statement, the first search result of each search resource can be recalled from a plurality of search resources associated with the target search business according to the target search statement, wherein the first search result is the recall result of the target search statement in each search resource, and the search business capability of the target search business in a plurality of search resources can be reflected through the first search result, so as to assist the R&D personnel of the target search business to timely and accurately understand the search business capability of the target search business.

[0096] Specifically, as shown in Figure 12 After obtaining the target search statement, the search background can be called to search the target search statement in a plurality of search resources associated with the target search business to obtain the first search result of each search resource. Specifically, a software development kit (SDK) can be placed in each search module corresponding to the search resource of the search background, and when the target search statement is obtained, the target search statement can be vertically searched through the calling link of the search background of the SDK of each search module corresponding to the search resource, and the search result obtained is the first search result.

[0097] For example, assuming that a target search business is a search applet S, the target search statement corresponding to the search applet S is "exhibition of B series cars of car brand A", and a plurality of search resources associated with the target search business are articles, applets, public numbers, expressions, services, and video numbers, etc. The search background can be called to vertically search the target search statement through the calling link of the search background of the SDK of each search module corresponding to the search resource, and the first search result of the target search statement in each search resource can be obtained, such as articles, applets, public numbers, expressions, services, and video numbers in the automobile industry.

[0098] In step S103, the first search result of each search resource is sorted to obtain the second search result of each search resource.

[0099] In this embodiment, after obtaining the first search result of each search resource, the first search result of each search resource can be sorted to obtain a better and more accurate second search result of each search resource, so that the subsequent diagnosis information of each search stage can be more accurately obtained according to the obtained second search result of each search resource, thereby the accuracy of obtaining the search result and the diagnosis information can be improved to a certain extent.

[0100] Specifically, after obtaining the first search result of each search resource, the first search result of each search resource can be sorted to obtain the second search result of each search resource. Specifically, the first search result of each search resource can be sorted according to the sorting rule corresponding to each search resource, such as mixed arrangement. For example, each recall result in the first search result of each search resource is scored based on a scoring strategy such as a similarity algorithm, and the second search result is sorted in descending order of similarity to the target search statement. Other sorting methods can also be used to sort the first search result of each search resource, which is not limited here.

[0101] In step S104, according to the second search result of each search resource, the processing of each search resource in each stage is diagnosed to obtain the diagnosis information of the target search service, wherein the diagnosis information includes the first diagnosis information for the recall stage and the second diagnosis information for the sorting stage.

[0102] In this embodiment, after obtaining the second search result of each search resource, the processing of each search resource in each stage can be diagnosed to obtain the first diagnosis information corresponding to the recall of the target search statement in the recall stage and the second diagnosis information corresponding to the sorting of the first search result in the sorting stage, so as to obtain the diagnosis information of the target search service.

[0103] The diagnosis information of the target search service for the target search statement is a set of the first diagnosis information corresponding to the recall of the target search statement in the recall stage and the second diagnosis information corresponding to the sorting of the first search result in the sorting stage, which can be used to reflect the specific search situation and defects of the target search service for the target search statement. The first diagnosis information is a detailed record of the processing of each search resource in the recall stage when recalling the target search statement. The second diagnosis information is a detailed record of the processing of each search resource in the sorting stage when sorting the first search result.

[0104] Specifically, as Figure 12As shown, after obtaining the second search result of each search resource, the processing of each search resource in the plurality of search resources at each stage can be diagnosed to obtain the diagnosis information of the target search service. Specifically, diagnosis can be performed when each search resource in the plurality of search resources recalls the target search statement at the recall stage. The algorithm features used when the first search result is recalled and the recall reasons of each result in the first search result, i.e., the first diagnosis information, can be obtained. Similarly, diagnosis can be performed when each search resource in the plurality of search resources sorts the first search result at the sorting stage. The algorithm features used when the first search result is sorted and the sorting reasons of the second search result, i.e., the second diagnosis information, can be obtained. Further, after obtaining the first diagnosis information of each search resource for the recall stage and the second diagnosis information of each search resource for the sorting stage, the information set obtained by aggregating the first diagnosis information of each search resource for the recall stage and the second diagnosis information of each search resource for the sorting stage, i.e., the diagnosis information of the target search service, can be obtained.

[0105] In the embodiments of the present application, a diagnosis method of a search service is provided. In the above manner, the first search result of each search resource can be recalled from a plurality of search resources associated with the target search service based on the target search statement. The influence of factors such as missing information and extraction errors in the search log can be avoided. Feature mining and diagnosis do not need to be repeatedly performed on the search log of the plurality of search resources. The second search result can be obtained based on the first search result of each search resource recalled at one time. The first diagnosis information for the recall stage and the second diagnosis information for the sorting stage can be directly diagnosed based on the second search result. Therefore, the accuracy and efficiency of the search result and the diagnosis information are improved.

[0106] Optionally, in the above Figure 2 Based on the corresponding embodiments, the diagnosis method of the search service provided in the embodiments of the present application includes another optional embodiment. As shown in the above Figure 3 The first search result of each search resource is recalled from a plurality of search resources based on the target search statement, and includes:

[0107] In step S201, the target search statement is subjected to word segmentation processing to obtain at least two to-be-processed words.

[0108] In step S202, the semantic recognition probability corresponding to each to-be-processed word is obtained by performing a semantic recognition operation on the at least two to-be-processed words.

[0109] In step S203, the industry category of the target search statement is determined based on the semantic recognition probability corresponding to each to-be-processed word.

[0110] In step S204, according to the industry category of the target search statement, the first search result of the vertical search of the target search statement in each search resource is obtained.

[0111] In this embodiment, after obtaining the target search statement, the target search statement can be subjected to word segmentation processing to obtain at least two to-be-processed words, and the semantic recognition operation is performed on the at least two to-be-processed words respectively to obtain the semantic recognition probability corresponding to each to-be-processed word. Then, the industry category of the target search statement can be determined according to the semantic recognition probability corresponding to each to-be-processed word, and the target search statement can be subjected to industry vertical search in each search resource according to the industry category of the target search statement to obtain the first search result of the vertical search of the target search statement in each search resource. The industry category to which the target search statement belongs can be determined, and the target search statement can be subjected to industry vertical search in each search resource, so that the massive unstructured information in each search resource can be extracted into structured statements similar or identical to the target search statement, i.e., the first search result, and the first search result can be recalled more accurately.

[0112] Specifically, after obtaining the target search statement, the target search statement can be subjected to word segmentation processing to obtain at least two to-be-processed words. Specifically, the target search statement obtained can be subjected to word segmentation processing by a Chinese word segmentation algorithm or an IK word segmentation algorithm or an algorithm based on a data dictionary, or other word segmentation algorithms, which are not limited here.

[0113] Further, after obtaining the at least two to-be-processed words, the semantic recognition operation can be performed on the at least two to-be-processed words respectively to obtain the semantic recognition probability corresponding to each to-be-processed word. Specifically, the at least two to-be-processed words can be input into a semantic recognition model, and the semantic recognition operation is performed on the to-be-processed words by the semantic recognition model to input the semantic recognition probability corresponding to each to-be-processed word. Other models can also be used to perform semantic recognition on the to-be-processed words, which are not limited here.

[0114] Further, after obtaining the semantic recognition probability of each to-be-processed word, the industry category of the target search statement can be determined according to the semantic recognition probability corresponding to each to-be-processed word. Specifically, the mutual information calculation can be performed on the semantic recognition probability, and the category corresponding to the calculated mutual information is determined as the industry category of the target search statement. Alternatively, the semantic recognition probability corresponding to each to-be-processed word can be calculated based on the Bayes formula, and the category corresponding to the maximum occurrence probability calculated is determined as the industry category of the target search statement. Other algorithms can also be used to determine the industry category of the target search statement, which are not limited here.

[0115] Further, after obtaining the industry category of the target search statement, the target search statement can be vertically searched in each search resource according to the industry category of the target search statement to obtain the first search result of the target search statement in each search resource. For example, assuming that a target search business is to search for a small program S and the corresponding target search statement is "exhibition of B series cars of car brand A", and the search resources associated with the target search business include articles, small programs, public accounts, emoticons, services, and video numbers, and the like, and assuming that the industry category of the target search statement is the automobile industry, the target search statement can be vertically searched in each search resource according to the industry category of the target search statement to obtain the first search result of the target search statement in each search resource, such as articles, small programs, public accounts, emoticons, services, and video numbers related to the words of cars, brand A, B series, or B series cars in the automobile industry.

[0116] Optionally, in the above Figure 2 Based on the corresponding embodiments, the search business diagnosis method provided by the embodiments of the present application further includes another optional embodiment, as shown in Figure 3 The first search result of each search resource is sorted to obtain the second search result of each search resource, including:

[0117] In step S301, the resource database is read to obtain the scoring strategy corresponding to each search resource;

[0118] In step S302, the first search result of each search resource is scored according to the scoring strategy to obtain the recall score of each search result in the first search result of each search resource;

[0119] In step S303, the first search result of each search resource is sorted according to the recall score of each search result to obtain the second search result of each search resource.

[0120] In this embodiment, after obtaining the first search result of each search resource, the resource database is read to obtain the scoring strategy corresponding to each search resource, and the first search result of each search resource is scored according to the scoring strategy to obtain the recall score of each search result in the first search result of each search resource. Then, the first search result of each search resource can be sorted according to the recall score of each search result to obtain the second search result of each search resource. The recall score of each search result in the first search result of each search resource can be accurately obtained through the scoring strategy, and the similarity between each search result and the target search statement can be numerically expressed through the recall score of each search result, so that the second search result most matched with the target search statement can be better and more accurately recalled.

[0121] Specifically, after obtaining the first search result of each search resource, it can be understood that before scoring the first search result of each search resource according to the scoring strategy, the filtering strategy corresponding to each search resource can be read from the resource database, and the search results in the first search result of each search resource can be filtered according to the filtering index or clustering algorithm in different filtering strategies, or other filtering methods, which are not limited here. The intermediate search result of each search resource after filtering can be obtained. Through the filtering strategy, the search results with a far semantic distance from the target search statement or the search results with a low similarity to the target search statement can be filtered, so that the second search result that best matches the target search statement can be recalled more accurately.

[0122] Further, after obtaining the intermediate search result of each search resource after filtering, the scoring strategy corresponding to each search resource can be read from the resource database and the first search result of each search resource can be scored according to the scoring strategy. Specifically, the Euclidean distance or cosine distance similarity algorithm can be used, or other scoring strategies such as a preset scoring index-value corresponding relationship algorithm can be used to score the intermediate search result of each search resource after filtering, so as to obtain the recall score of each search result in the intermediate search result of each search resource after filtering. Other scoring methods can also be used, which are not limited here.

[0123] Further, if the cosine distance is used to calculate the recall score of each search result in the intermediate search result of each search resource after filtering, since the recall score is positively correlated with the similarity, the intermediate search result of each search resource after filtering can be sorted in descending order of the recall score of each search result, so as to obtain the second search result of each search resource. Similarly, if the recall score is negatively correlated with the similarity, the intermediate search result of each search resource after filtering can be sorted in ascending order of the recall score of each search result, so as to obtain the second search result of each search resource.

[0124] Optionally, in the above Figure 2 On the basis of the corresponding embodiment, the search service diagnosis method provided by the embodiment of the present application further includes another optional embodiment, as shown in Figure 4 The target search statement carries a synchronous reporting parameter. After obtaining the second search result of each search resource, the processing of each search resource in each stage is diagnosed, and the diagnosis information of the target search service is obtained. The method further includes:

[0125] In step S401, a synchronous reporting path is acquired according to a synchronous reporting parameter, wherein the synchronous reporting parameter is used to control the synchronous reporting switch to open the synchronous reporting path.

[0126] In step S402, the first search result, the second search result and the diagnosis information of the target search service of each search resource are uploaded to the diagnosis database through the synchronous reporting path.

[0127] In the embodiment, after the diagnosis information of the target search service is acquired, the synchronous reporting switch can be controlled to open the synchronous reporting path through the synchronous reporting parameter, and the synchronous reporting path can be acquired according to the synchronous reporting parameter, and then the first search result, the second search result and the diagnosis information of the target search service of each search resource can be accurately uploaded to the diagnosis database through the synchronous reporting path, avoiding data loss, so that the accuracy and integrity of the acquired search result and diagnosis result can be improved to a certain extent.

[0128] Specifically, as shown in Figure 12 After the target search statement is acquired, if the acquired target search statement carries the synchronous reporting parameter, the synchronous reporting switch can be controlled to open the synchronous reporting path through the synchronous reporting parameter, wherein the synchronous reporting parameter can be in the form of a synchronization number, a synchronization label or a synchronization symbol, and can also be in other parameter forms, which are not limited here.

[0129] Further, after the synchronous reporting switch opens the synchronous reporting path, the synchronous reporting path can be accurately acquired through the synchronous reporting parameter, so that the first search result, the second search result and the diagnosis information of the target search service of each search resource acquired can be timely and accurately uploaded to the diagnosis database for storage, so that the researchers of the target search service can real-time retrieve the first search result, the second search result and the diagnosis information of the target search service of each search resource reported through the synchronous reporting path from the diagnosis database.

[0130] It can be understood that when it is not necessary to locate the business problem of the target search service, the synchronous reporting switch can be in the closed state, and only the acquired target search statement is searched to acquire the first search result and the second search result, without diagnosing each search stage, so that the influence of the reporting of the diagnosis information on the target search service itself can be reduced.

[0131] Optionally, based on the above Figure 4 corresponding embodiment, another optional embodiment of the diagnosis method of the search service provided by the embodiment of the present application is as follows: Figure 5As shown, before uploading the first search result, the second search result and the diagnostic information of the target search service of each search resource to the diagnostic database through the synchronous reporting path, the method further comprises: step S501; step S402 comprises: step S502.

[0132] In step S501, the first search result, the second search result and the diagnostic information of the target search service of each search resource are standardized respectively to obtain the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource.

[0133] In step S502, the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource are uploaded to the diagnostic database through the synchronous reporting path.

[0134] In this embodiment, before uploading the first search result, the second search result and the diagnostic information of the target search service of each search resource to the diagnostic database through the synchronous reporting path, the first search result, the second search result and the diagnostic information of the target search service of each search resource can be standardized respectively to obtain the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource, and then the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource can be uploaded to the diagnostic database through the synchronous reporting path. By standardizing the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource, the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource with unified format can be obtained, which can facilitate the system to manage the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource and meet the personalized display requirements of users, etc.

[0135] Specifically, after obtaining the first search result, the second search result and the diagnostic information of the target search service of each search resource, the first search result, the second search result and the diagnostic information of the target search service of each search resource can be standardized respectively to obtain the first standard search result, the second standard search result and the standard diagnostic information of the target search service of each search resource. Specifically, the data can be standardized by customizing key-value pairs such as key value, for example, the first search result, the second search result and the diagnostic information of the target search service of each search resource can be converted into a standardized format according to the key-value pair format of name: xxx and phone: xxx.

[0136] It can be understood that, due to the differences in SDK call links and call chain lengths of search modules corresponding to different search resources, the diagnosis information is complex and diverse, therefore, the resource number can be configured in the SDK of the search module corresponding to the search resource to distinguish the categories of different search resources, and the search logic is divided into search stages, for example, the search stages include L1-L5, wherein L1 and L2 represent the recall stage, L3-L4 represent the precision stage, i.e., the filtering stage, and L5 represents the mixed arrangement, i.e., the sorting stage, or L1-L4 represent the vertical search stage, and L5 represents the mixed arrangement, i.e., the sorting stage, and other search stage division is also possible, which is not limited specifically herein. Meanwhile, the SDK of the search module corresponding to the search resource can also be configured with a custom sub-stage, i.e., the first search result, the second search result, and the diagnosis information of the target search business of each search resource are standardized by a custom key-value pair, and it can be understood that these configuration information can be configured before each SDK obtains the first search result, the second search result, and the diagnosis information of the target search business of each search resource, and can be organized in the std::unordered_map in the Libco coroutine variable maintained by the SDK, thereby achieving the purpose of reuse. Meanwhile, the SDK can add a coroutine lock (Libco::Coutex) when obtaining the first search result, the second search result, the diagnosis information of the target search business, and reading the information, so as to ensure thread safety, thereby maintaining the safety of obtaining the search result and the diagnosis result to a certain extent.

[0137] Further, after obtaining the first standard search result, the second standard search result, and the standard diagnosis information of the target search business of each search resource, the first standard search result, the second standard search result, and the standard diagnosis information of the target search business of each search resource can be uploaded to the diagnosis database through a synchronous reporting path.

[0138] Optionally, in the above Figure 4 Based on the corresponding embodiments, another optional embodiment of the diagnosis method of the search business provided by the embodiments of the present application is as shown in Figure 6 The first search result, the second search result, and the diagnosis information of the target search business of each search resource are uploaded to the diagnosis database through a synchronous reporting path, and the method comprises the steps of:

[0139] In step S601, the first search result, the second search result, and the diagnosis information of the target search business of each search resource are written into a target request packet to obtain a target request packet;

[0140] In step S602, the target request packet is uploaded to the diagnosis database through a synchronous reporting path.

[0141] In this embodiment, after obtaining the first search result, the second search result, and the diagnostic information of the target search service for each search resource, the first search result, the second search result, and the diagnostic information of the target search service for each search resource can be written into a request response packet to obtain a target request response packet. Then, the target request response packet can be uploaded to the diagnostic database through the synchronous reporting path. By writing the first search result, the second search result, and the diagnostic information of the target search service for each search resource layer by layer, the request response packet can carry the written first search result, the second search result, and the diagnostic information of the target search service for each search resource layer by layer, so that the data can be uploaded to the diagnostic database securely and accurately. The search results and diagnostic results can be uploaded to the diagnostic database at once through the request response packet, thereby maintaining the accuracy, completeness, and efficiency of obtaining search results and diagnostic results to a certain extent.

[0142] Specifically, such as Figure 12 As shown, after obtaining the first search result, the second search result, and the diagnostic information of the target search business for each search resource, the first search result, the second search result, and the diagnostic information of the target search business for each search resource are written layer by layer into the request response packet. This allows the request response packet to carry the first search result, the second search result, and the diagnostic information of the target search business for each search resource layer by layer, thereby enabling the target request response packet to be uploaded to the diagnostic database through the synchronous reporting path.

[0143] Understandably, to avoid excessively large request response packets, the SDK also supports write storage. Specifically, it can store the first and second search results of each search resource, as well as the diagnostic information of the target search business, in the SDK's resource database. When it is necessary to report the first and second search results of each search resource and the diagnostic information of the target search business, the SDK can retrieve the first and second search results of each search resource and the diagnostic information of the target search business from the resource database and upload them to the diagnostic database through the synchronous reporting path.

[0144] It can be understood that after obtaining the first search result, the second search result and the diagnosis information of the target search service of each search resource, the background module can call the SDK reporting interface of the synchronous reporting path for reporting. The first search result, the second search result and the diagnosis information of the target search service of each search resource can be reported in the form of a document or a document set. The SDK reporting interface supports reporting of document recall, filtering, features and other information. The recall state is divided into manual intervention, normal recall and the like. The filtering state is divided into business strategy filtering, sorting filtering, blocking and the like. The feature information is reported in the form of fixed public feature fields and custom key-value pairs. The reporting format is configured by the search resource. It can be understood that the reporting interface can use Libco coroutine variables and fixed-size arrays to ensure performance and thread safety. Specifically, the reporting array size can be limited to avoid reporting too much data and causing the entire system to be overwhelmed. However, some data may be slightly larger than the set size. If the data volume of the first search result, the second search result and the diagnosis information of the target search service of each search resource exceeds the preset array size, the array can be expanded when the SDK object is destructed. This can avoid reporting too much data due to unintentional bugs, so that a stable state can be reached after the service runs for a period of time, thereby maintaining the safe reporting of data.

[0145] Optionally, in the above Figure 4 Based on the corresponding embodiments, the diagnosis method of the search service provided by the embodiments of the present application further includes another optional embodiment, as shown in Figure 7 After uploading the first search result, the second search result and the diagnosis information of the target search service of each search resource to the diagnosis database through the synchronous reporting path, the method further includes:

[0146] In step S701, the first search result, the second search result and the diagnosis information of the target search service of each search resource are parsed by a syntax tree to obtain the first parsing result, the second parsing result and the diagnosis parsing information of the target search service of each search resource.

[0147] In step S702, a search diagnosis list of the target search service is generated according to the first parsing result, the second parsing result and the diagnosis parsing information of the target search service of each search resource according to a list generation template.

[0148] In this embodiment, after uploading the first and second search results of each search resource and the diagnostic information of the target search service to the diagnostic database via a synchronous reporting path, syntax tree parsing can be performed on the first and second search results of each search resource and the diagnostic information of the target search service to obtain the first and second parsing results of each search resource and the diagnostic parsing information of the target search service. Then, a search diagnostic list for the target search service can be generated according to a list generation template based on the first and second parsing results of each search resource and the diagnostic parsing information of the target search service. The syntax tree parsing allows for depth-first traversal of the data to ensure complete reading of the first and second search results of each search resource and the diagnostic information of the target search service in the diagnostic database. Furthermore, the parsed first and second parsing results of each search resource and the diagnostic parsing information of the target search service can be standardized to obtain a search diagnostic list for the target search service that can be displayed in a unified format. This facilitates the maintenance and management of the search results and diagnostic results of the target search service, avoids data loss, and improves the accuracy and completeness of the obtained search results and diagnostic results to a certain extent.

[0149] Specifically, such as Figure 12 As shown, after uploading the first and second search results of each search resource and the diagnostic information of the target search business to the diagnostic database via a synchronous reporting path, a syntax tree parsing can be performed on the first and second search results of each search resource and the diagnostic information of the target search business. This allows the obtained first and second search results of each search resource and the diagnostic information of the target search business to be expressed in a unified format or style, obtaining the first and second parsing results of each search resource and the diagnostic parsing information of the target search business. Then, based on the first and second parsing results of each search resource and the diagnostic parsing information of the target search business, a search diagnostic list for the target search business can be generated according to a list generation template. This allows for quick indexing of the search diagnostic list of the target search business through search identifiers and a one-time push to the R&D personnel. The list generation template includes list styles and list generation rules. For example, the first and second parsing results of each search resource and the diagnostic parsing information of the target search business can be subdivided into document information, document set information, user information, query information, vertical search information, and module information corresponding to each search resource through the list generation template.

[0150] It can be understood that, in the reporting mode of the self-defined key-value pair, the same format of the search diagnosis list is generated by combining the list generation template, which not only meets the requirement of regular and unified display of the search diagnosis list, but also meets the requirement of individual customization of different search resource modules, so that the R&D personnel can more easily find the missing information of the target search service through the search diagnosis list, and the performance of the target search service can be improved in time, and the search results and diagnosis results of each search resource are no longer isolated, and data missing is avoided.

[0151] It can be understood that, after obtaining the search diagnosis list, the obtained search diagnosis list can be decomposed and regularized and stored in a database, memory or the like, and the stored search diagnosis list is used as the data cache of the front-end interaction of the diagnosis system. The cache data needs to be cleaned regularly to avoid occupying too much storage.

[0152] Optionally, in the above Figure 2 Based on the corresponding embodiment, another optional embodiment of the diagnosis method of the search service provided by the embodiment of the application is as shown in Figure 8 The target search statement carries an asynchronous reporting parameter, the processing of each search resource in each stage is diagnosed according to the second search result of each search resource, and after the diagnosis information of the target search service is obtained, the method further includes:

[0153] In step S801, the asynchronous reporting path is obtained according to the asynchronous reporting parameter, wherein the asynchronous reporting parameter is used to control the asynchronous reporting switch to open the asynchronous reporting path.

[0154] In step S802, the first search result, the second search result of each search resource and the diagnosis information of the target search service are uploaded to the diagnosis database through the asynchronous reporting path.

[0155] In this embodiment, the diagnosis information of the target search service can be obtained by controlling the asynchronous reporting switch to open the asynchronous reporting path through the asynchronous reporting parameter, and the asynchronous reporting path is obtained according to the asynchronous reporting parameter. Then, the first search result, the second search result of each search resource and the diagnosis information of the target search service can be accurately and quickly uploaded to the diagnosis database through the asynchronous reporting path, which can avoid real-time uploading and reduce resource consumption. The data is more quickly stored to the diagnosis database for R&D personnel to call at any time in the future, so as to improve the accuracy and efficiency of obtaining the search result and the diagnosis result to a certain extent.

[0156] Specifically, as Figure 12As shown, after the target search statement is acquired, if the acquired target search statement carries an asynchronous reporting parameter, the asynchronous reporting path can be started by the asynchronous reporting parameter controlling the asynchronous reporting switch, wherein the asynchronous reporting parameter can be in the form of an asynchronous number, an asynchronous label or an asynchronous symbol, and can also be in other parameter forms, which are not limited here.

[0157] Further, after the asynchronous reporting path is started by the asynchronous reporting switch, the asynchronous reporting path can be accurately acquired by the asynchronous reporting parameter, so that the first search result, the second search result and the diagnostic information of the target search service of each search resource acquired can be timely and accurately uploaded to the diagnostic database for storage through the asynchronous reporting path, so that the subsequent target search service researchers can retrieve the first search result, the second search result and the diagnostic information of the target search service of each search resource reported through the asynchronous reporting path from the diagnostic database at any time in the future.

[0158] It can be understood that when it is unnecessary to locate the business problem of the target search service, i.e., it is not necessary to locate the business problem of the target search service in the future, the asynchronous reporting switch can be in the closed state, and only the acquired target search statement is searched to acquire the first search result and the second search result, without diagnosing each search stage, so that the influence of the reporting of the diagnostic information on the target search service itself can be reduced.

[0159] It can be understood that when it is necessary to locate the business problem of the target search service, i.e., it is necessary to locate the business problem of the target search service at present or in the future, the synchronous reporting switch and the asynchronous reporting switch can be in the open state, not only the acquired target search statement can be searched to acquire the first search result and the second search result, but also each search stage can be diagnosed to comprehensively acquire the diagnostic information of each search stage, so that the synchronous diagnosis and the historical diagnosis can be performed at the same time, which can be understood as the request generated during the synchronous diagnosis can also be traced back, the reuse rate of the search result and the diagnostic result can be improved, the workload of the repeated search and diagnosis of the researchers can be reduced, the vacancy of the historical search result and the diagnostic result can be made up, and a better information recording and problem locating mode can be achieved. It can be understood that the asynchronous diagnostic information and the synchronous diagnostic information are completely consistent.

[0160] Optionally, in the above Figure 8 Based on the corresponding embodiment, the search service diagnosis method provided by the embodiment of the present application further includes another optional embodiment, which is as shown in Figure 9 The first search result, the second search result and the diagnostic information of the target search service of each search resource are uploaded to the diagnostic database through the asynchronous reporting path, including:

[0161] In step S901, the first search result, the second search result and the diagnosis information of the target search service of each search resource are summarized to obtain a diagnosis data set of the target search service;

[0162] In step S902, the diagnosis data set of the target search service is written into the first message queue through an asynchronous reporting path;

[0163] In step S903, the diagnosis data set of the target search service in the first message queue is aggregated according to a time dimension to obtain aggregated diagnosis data corresponding to each time window of the target search service;

[0164] In step S904, the aggregated diagnosis data corresponding to each time window of the target search service is uploaded to the diagnosis database through the asynchronous reporting path.

[0165] In this embodiment, after obtaining the first search result, the second search result and the diagnosis information of the target search service of each search resource, the first search result, the second search result and the diagnosis information of the target search service of each search resource can be summarized to obtain a diagnosis data set of the target search service, and the diagnosis data set of the target search service is written into the first message queue through an asynchronous reporting path. The diagnosis data set of the target search service in the first message queue can be aggregated according to a time dimension to obtain aggregated diagnosis data corresponding to each time window of the target search service. Then, the aggregated diagnosis data corresponding to each time window of the target search service can be uploaded to the diagnosis database through the asynchronous reporting path. The first search result, the second search result and the diagnosis information of the target search service of each search resource can be summarized, i.e., first combined, and written into the first message queue, which can avoid the performance impact caused by multiple information writing. In addition, the search results and diagnosis information belonging to the same target search statement in the first message queue can be aggregated together, which can avoid the loss of search results and diagnosis information, thereby improving the accuracy and integrity of the obtained search results and diagnosis results to a certain extent.

[0166] Specifically, as Figure 12As shown, after obtaining the first search result, the second search result, and the diagnostic information of the target search service of each search resource, the first search result, the second search result, and the diagnostic information of the target search service of each search resource can be summarized to obtain a diagnostic data set of the target search service, and the diagnostic data set of the target search service is written into the first message queue through an asynchronous reporting path. The diagnostic data set of the target search service in the first message queue can be aggregated according to a time dimension to obtain aggregated diagnostic data corresponding to each time window of the target search service. Specifically, the obtained diagnostic data set of the target search service can be written into a disk-based queue, i.e., the first message queue, for peak shaving and backtracking.

[0167] Further, the first message queue can be used to consume data, and the data can be aggregated in memory based on a time window period, such as a 5s window. All search results and diagnostic results belonging to the same target search statement can be aggregated together. It can be understood that the time window can be calculated based on a timestamp in the diagnostic information, which is a non-system processing time. Then, the obtained aggregated diagnostic data corresponding to each time window of the target search service can be uploaded to a diagnostic database through an asynchronous reporting path for storage. In this way, within the data storage time, a developer can read the aggregated diagnostic data corresponding to the search identifier at any time after the data is stored.

[0168] It can be understood that after the aggregated diagnostic data corresponding to each time window of the target search service is uploaded to the diagnostic database through the asynchronous reporting path for storage, the aggregated diagnostic data in the memory can be written into object storage. Specifically, the data can be indexed and written into LSM-based storage. For example, there are tens of millions of rows of data in the data, including student numbers, names, and scores. The student numbers can be indexed with other data, so that the indexed data can be searched by the student number in milliseconds in the future, thereby improving the efficiency of obtaining search results and diagnostic results to a certain extent.

[0169] Optionally, in the above Figure 9 Based on the corresponding embodiments, in another optional embodiment of the diagnostic method of the search service provided by the embodiments of the present application, as shown in Figure 10 Before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, the method further includes steps S1001 and S1002, and step S902 includes step S1003.

[0170] In step S1001, the current resource consumption corresponding to the target search service is obtained.

[0171] In step S1002, if the current resource consumption corresponding to the target search service is greater than or equal to the resource threshold, the diagnostic data set of the target search service is compressed to obtain diagnostic compressed data of the target search service.

[0172] In step S1003, the diagnostic compressed data of the target search service is written into the first message queue through the asynchronous reporting path.

[0173] In this embodiment, before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, the current resource consumption corresponding to the target search service can be obtained. When the current resource consumption corresponding to the target search service is greater than or equal to the resource threshold, the diagnostic data set of the target search service can be compressed to obtain diagnostic compressed data of the target search service, and then the diagnostic compressed data of the target search service can be written into the first message queue through the asynchronous reporting path. Whether the diagnostic data set of the target search service needs to be compressed and then uploaded can be determined according to the current resource consumption, so as to reduce resource consumption.

[0174] Specifically, before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, the main function of the search background module is to provide search services for users, which will consume CPU memory. Therefore, if too many resources of the machine are occupied before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, online failure will be caused. Therefore, the current resource consumption corresponding to the target search service can be obtained. If the current resource consumption corresponding to the target search service is greater than or equal to the resource threshold, it can be understood that too many resources of the machine are currently occupied. In order to reduce the consumption of CPU memory, the diagnostic data set of the target search service can be compressed to obtain diagnostic compressed data of the target search service, and then the diagnostic compressed data of the target search service can be written into the first message queue through the asynchronous reporting path by selecting network IO.

[0175] Optionally, in the above Figure 9 Based on the corresponding embodiments, in another optional embodiment of the search service diagnostic method provided by the embodiments of the present application, as shown in Figure 10 The target search statement carries a search identifier, and the first search result, the second search result and the diagnostic information of the target search service are summarized to obtain a diagnostic data set of the target search service, including:

[0176] In step S1101, target routing information of a target search service is obtained according to a search identifier, wherein the target routing information is obtained by hashing the search identifier.

[0177] In step S1102, the first search result, the second search result and the diagnosis information of the target search service of each search resource are routed to the target node corresponding to the target routing information for aggregation to obtain a diagnosis data set of the target search service.

[0178] In the embodiment, after obtaining the first search result, the second search result and the diagnosis information of the target search service of each search resource, the target routing information of the target search service obtained by hashing can be acquired according to the search identifier, and then the first search result, the second search result and the diagnosis information of the target search service of each search resource can be routed to the target node corresponding to the target routing information for aggregation to obtain a diagnosis data set of the target search service. The first search result, the second search result and the diagnosis information of the target search service of each search resource can be routed to the target node according to the target routing information for aggregation and then uploaded to the diagnosis database, which can avoid repeated routing of data and thus reduce repeated uploading of data, and can improve the efficiency of obtaining search results and diagnosis results to a certain extent.

[0179] Specifically, after obtaining the first search result, the second search result and the diagnosis information of the target search service of each search resource, the target routing information of the target search service obtained by hashing can be acquired according to the search identifier, which can be consistent hash routing through the search identifier such as a user account or a search request number, and then the target routing information of the target search service can be acquired. The first search result, the second search result and the diagnosis information of the target search service of each search resource belonging to the same target search statement pair can be routed to the target node such as the same machine according to the target routing information for aggregation to obtain a diagnosis data set of the target search service.

[0180] Optionally, in the above Figure 10 Based on the corresponding embodiment, the diagnosis method of the search service provided in the embodiment of the present application further includes another optional embodiment, as shown in Figure 11 Before the diagnosis data set of the target search service is written into the first message queue through the asynchronous reporting path, the method further includes steps S1201 and S1202, and step S902 includes step S1203.

[0181] In step S1201, the search identifier is indexed to the delay reporting data and the target node change data.

[0182] In step S1203, the delay reporting data and the target node change data are aggregated to obtain a delay data set.

[0183] In step S1203, the diagnostic data set of the target search service and the delay data set are written into the first message queue through the asynchronous reporting path.

[0184] In this embodiment, before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, the delay reporting data and the target node change data can be summarized according to the index between the search identifier and the delay reporting data and the target node change data to obtain the delay data set, and then the diagnostic data set of the target search service and the delay data set can be written into the first message queue through the asynchronous reporting path. The missing data can be supplemented by searching the delay reporting data and the target node change data, so that the delay reporting data and the target node change data are summarized and uploaded again, which can avoid information loss and improve the accuracy and integrity of the obtained search results and diagnostic results to a certain extent.

[0185] The delay reporting data can be any one of the first search result, the second search result, or the diagnostic information of the target search service, and the target node change data can be any one of the first search result, the second search result, or the diagnostic information of the target search service.

[0186] Specifically, after obtaining the first search result, the second search result, and the diagnostic information of the target search service of each search resource, the first search result, the second search result, and the diagnostic information of the target search service of each search resource can be affected by the network or resource load, resulting in a delay in reporting the first search result, the second search result, and the diagnostic information of the target search service of each search resource, or a change in the number of backend service nodes, or a change in the target node, etc. Therefore, according to the index relationship between the search identifier and the diagnostic data set, the delay reporting data and the target node change data can be quickly and accurately indexed, and the indexed delay reporting data and target node change data can be summarized to obtain the delay data set. Then, the diagnostic data set of the target search service and the delay data set can be written into the first message queue through the asynchronous reporting path.

[0187] The search service diagnosis device in the present application will be described in detail below. Please refer to Figure 13 , Figure 13 An embodiment of the search service diagnosis device in the present application is shown in the figure. The search service diagnosis device 20 comprises:

[0188] The acquisition unit 201 is configured to acquire a target search statement for a target search service.

[0189] The processing unit 202 is configured to call first search results of each search resource from a plurality of search resources according to a target search statement, wherein the plurality of search resources belong to search resources associated with the target search service;

[0190] The processing unit 202 is further configured to sort the first search results of each search resource to obtain second search results of each search resource;

[0191] The processing unit 202 is further configured to diagnose the processing of each search resource in each stage according to the second search results of each search resource, to obtain diagnosis information of the target search service, wherein the diagnosis information includes first diagnosis information for the calling stage and second diagnosis information for the sorting stage.

[0192] Optionally, in the above Figure 13 Based on the corresponding embodiment, the processing unit 202 in another embodiment of the diagnosis device for the search service provided by the present application can be specifically configured to:

[0193] Carrying out word segmentation processing on the target search statement to obtain at least two to-be-processed words;

[0194] Respectively performing semantic recognition operations on the at least two to-be-processed words to obtain semantic recognition probabilities corresponding to each to-be-processed word;

[0195] Determining an industry category of the target search statement according to the semantic recognition probabilities corresponding to each to-be-processed word;

[0196] According to the industry category of the target search statement, obtaining first search results of the target search statement in each search resource.

[0197] Optionally, in the above Figure 13 Based on the corresponding embodiment, the processing unit 202 in another embodiment of the diagnosis device for the search service provided by the present application can be specifically configured to:

[0198] Reading the resource database to obtain a scoring strategy corresponding to each search resource;

[0199] According to the scoring strategy, scoring the first search results of each search resource to obtain recall scores of each search result in the first search results of each search resource;

[0200] According to the recall scores of each search result, sorting the first search results of each search resource to obtain second search results of each search resource.

[0201] Optionally, in the above Figure 13 Based on the corresponding embodiment, the processing unit 202 in another embodiment of the diagnosis device for the search service provided by the present application can be specifically configured to:

[0202] The acquisition unit 201 is further configured to acquire a synchronous reporting path according to a synchronous reporting parameter, wherein the synchronous reporting parameter is used to control the synchronous reporting switch to open the synchronous reporting path.

[0203] The processing unit 202 is further configured to upload the first search result, the second search result and the diagnosis information of the target search service of each search resource to the diagnosis database through the synchronous reporting path.

[0204] Optionally, in the above Figure 13 Based on the corresponding embodiment, another embodiment of the search service diagnosis apparatus provided by the present application is provided, wherein the processing unit 202 is specifically configured to:

[0205] The processing unit 202 is further configured to perform standardization processing on the first search result, the second search result and the diagnosis information of the target search service of each search resource respectively, to obtain the first standard search result, the second standard search result and the standard diagnosis information of the target search service of each search resource.

[0206] The processing unit 202 is specifically configured to upload the first standard search result, the second standard search result and the standard diagnosis information of the target search service of each search resource to the diagnosis database through the synchronous reporting path.

[0207] Optionally, in the above Figure 13 Based on the corresponding embodiment, another embodiment of the search service diagnosis apparatus provided by the present application is provided, wherein the processing unit 202 is specifically configured to:

[0208] The processing unit 202 is specifically configured to write the first search result, the second search result and the diagnosis information of the target search service of each search resource into a target request back packet.

[0209] The processing unit 202 is specifically configured to upload the target request back packet to the diagnosis database through the synchronous reporting path.

[0210] Optionally, in the above Figure 13 Based on the corresponding embodiment, another embodiment of the search service diagnosis apparatus provided by the present application is provided, wherein the processing unit 202 is specifically configured to:

[0211] The processing unit 202 is further configured to perform syntax tree analysis on the first search result, the second search result and the diagnosis information of the target search service of each search resource, to obtain the first analysis result, the second analysis result and the diagnosis analysis information of the target search service of each search resource.

[0212] The generation unit 203 is configured to generate a search diagnosis list of the target search service according to the first analysis result, the second analysis result and the diagnosis analysis information of the target search service of each search resource according to a list generation template.

[0213] Optionally, in the above Figure 13 Based on the corresponding embodiments, another embodiment of the diagnosis device for search service provided by the embodiments of the present application is provided,

[0214] The acquisition unit 201 is further configured to acquire an asynchronous reporting path according to an asynchronous reporting parameter, wherein the asynchronous reporting parameter is used to control the asynchronous reporting switch to open the asynchronous reporting path.

[0215] The processing unit 202 is further configured to upload the first search result, the second search result and the diagnosis information of the target search service of each search resource to the diagnosis database through the asynchronous reporting path.

[0216] Optionally, in the above Figure 13 Based on the corresponding embodiments, the processing unit 202 in another embodiment of the diagnosis device for search service provided by the embodiments of the present application can be specifically configured to:

[0217] aggregate the first search result, the second search result and the diagnosis information of the target search service of each search resource to obtain a diagnosis data set of the target search service;

[0218] write the diagnosis data set of the target search service into the first message queue through the asynchronous reporting path;

[0219] perform aggregation processing on the diagnosis data set of the target search service in the first message queue according to a time dimension to obtain aggregation diagnosis data corresponding to each time window of the target search service;

[0220] upload the aggregation diagnosis data corresponding to each time window of the target search service to the diagnosis database through the asynchronous reporting path.

[0221] Optionally, in the above Figure 13 Based on the corresponding embodiments, another embodiment of the diagnosis device for search service provided by the embodiments of the present application is provided,

[0222] The acquisition unit 201 is further configured to acquire a current resource consumption of the target search service;

[0223] The processing unit 202 is further configured to perform data compression on the diagnosis data set of the target search service to obtain diagnosis compressed data of the target search service if the current resource consumption of the target search service is greater than or equal to the resource threshold.

[0224] The processing unit 202 can be specifically configured to write the diagnosis compressed data of the target search service into the first message queue through the asynchronous reporting path.

[0225] Optionally, in the above Figure 13Based on the corresponding embodiments, in another embodiment of the search service diagnosis apparatus provided by the present application, the processing unit 202 can be specifically configured to:

[0226] obtain target routing information of the target search service according to the search identifier, wherein the target routing information is obtained by hashing the search identifier;

[0227] route the first search result, the second search result and the diagnosis information of the target search service of each search resource to the target node corresponding to the target routing information for aggregation, to obtain a diagnosis data set of the target search service.

[0228] Optionally, in the above Figure 13 Based on the corresponding embodiments, in another embodiment of the search service diagnosis apparatus provided by the present application,

[0229] The processing unit 202 is further configured to index the delay reporting data and the target node change data according to the search identifier.

[0230] The processing unit 202 is further configured to aggregate the delay reporting data and the target node change data to obtain a delay data set.

[0231] The processing unit 202 can be specifically configured to: write the diagnosis data set and the delay data set of the target search service to the first message queue through an asynchronous reporting path.

[0232] Another aspect of the present application provides another computer device schematic diagram, as shown in Figure 14 , and Figure 14 is a computer device structure schematic diagram provided by an embodiment of the present application. The computer device 300 can have great differences due to different configurations or performances, and can include one or more central processing units (CPU) 310 (for example, one or more processors) and a memory 320, one or more storage media 330 (for example, one or more mass storage devices) storing application programs 331 or data 332. The memory 320 and the storage medium 330 can be temporary storage or persistent storage. The programs stored in the storage medium 330 can include one or more modules (not shown in the figure), and each module can include a series of instruction operations in the computer device 300. Further, the central processing unit 310 can be configured to communicate with the storage medium 330 and execute a series of instruction operations in the storage medium 330 on the computer device 300.

[0233] The computer device 300 can also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 333, such as Windows Server TM , Mac OS X TM , Unix TM , Linux TM , FreeBSD TM , and the like.

[0234] The computer device 300 described above is further configured to perform the steps of any of the embodiments described above. Figures 2 to 11

[0235] Another aspect of the present application provides a computer readable storage medium having stored therein instructions which, when executed on a computer, cause the computer to perform the steps of any of the methods described in the embodiments described above. Figures 2 to 11

[0236] Another aspect of the present application provides a computer program product containing instructions which, when executed on a computer or processor, cause the computer or processor to perform the steps of any of the methods described in the embodiments described above. Figures 2 to 11

[0237] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0238] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented by other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0239] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0240] ​​​In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0241] When the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application, essentially or in the form of a contribution to the prior art, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

Claims

1. A diagnostic method of a search service, characterized by, The method comprises the following steps: When search service diagnosis is required, a target search statement for a target search service is obtained; the target search statement carries a synchronous reporting parameter and an asynchronous reporting parameter; According to the target search statement, a first search result of each search resource is recalled from a plurality of search resources, wherein the plurality of search resources belong to search resources associated with the target search service; The first search result of each search resource is sorted in a mixed arrangement manner to obtain a second search result of each search resource; According to the second search result of each search resource, the processing of each search resource in each stage is diagnosed to obtain diagnosis information of the target search service, wherein the diagnosis information includes first diagnosis information corresponding to the recall of the target search statement by each search resource in the recall stage and second diagnosis information corresponding to the sorting of the first search result by each search resource in the sorting stage; According to the synchronous reporting parameter, a synchronous reporting path is obtained, wherein the synchronous reporting parameter is used to control the synchronous reporting switch to open the synchronous reporting path; The first search result, the second search result of each search resource, and the diagnosis information of the target search service are standardized by using a self-defined key-value pair to obtain the first standard search result, the second standard search result of each search resource, and the standard diagnosis information of the target search service; The first standard search result, the second standard search result of each search resource, and the standard diagnosis information of the target search service are uploaded to a diagnosis database through the synchronous reporting path; According to the asynchronous reporting parameter, an asynchronous reporting path is obtained, wherein the asynchronous reporting parameter is used to control the asynchronous reporting switch to open the asynchronous reporting path; The first search result, the second search result of each search resource, and the diagnosis information of the target search service are uploaded to the diagnosis database through the asynchronous reporting path; After the first standard search result, the second standard search result of each search resource, and the standard diagnosis information of the target search service are uploaded to the diagnosis database through the synchronous reporting path, the method further comprises: The first standard search result, the second standard search result of each search resource, and the standard diagnosis information of the target search service are parsed by a syntax tree to obtain the first parsing result, the second parsing result of each search resource, and the diagnosis parsing information of the target search service; According to the first parsing result, the second parsing result of each search resource, and the diagnosis parsing information of the target search service, a search diagnosis list of the target search service is generated according to a list generation template.

2. The method of claim 1, wherein, According to the target search statement, a first search result of each search resource is recalled from a plurality of search resources, wherein the plurality of search resources belong to search resources associated with the target search service; The target search statement is processed by word segmentation to obtain at least two to-be-processed words; Each to-be-processed word is subjected to a semantic recognition operation to obtain a semantic recognition probability corresponding to each to-be-processed word; Determine the industry category of the target search statement according to the semantic recognition probability corresponding to each of the to-be-processed words; According to the industry category of the target search statement, the first search result of the target search statement in each search resource is obtained.

3. The method of claim 1, wherein, The first search result of each search resource is sorted to obtain the second search result of each search resource, including: Reading a resource database to obtain a scoring strategy corresponding to each search resource; According to the scoring strategy, the first search result of each search resource is scored to obtain a recall score of each search result in the first search result of each search resource; According to the size of the recall score of each search result, the first search result of each search resource is sorted to obtain the second search result of each search resource.

4. The method of claim 1, wherein, The first standard search result of each search resource, the second standard search result, and the standard diagnostic information of the target search service are uploaded to the diagnostic database through the synchronous reporting path, including: The first standard search result of each search resource, the second standard search result, and the standard diagnostic information of the target search service are written into a target request packet; The target request packet is uploaded to the diagnostic database through the synchronous reporting path.

5. The method of claim 1, wherein, The first search result of each search resource, the second search result, and the diagnostic information of the target search service are uploaded to the diagnostic database through the asynchronous reporting path, including: The first search result of each search resource, the second search result, and the diagnostic information of the target search service are summarized to obtain a diagnostic data set of the target search service; The diagnostic data set of the target search service is written into a first message queue through the asynchronous reporting path; The diagnostic data set of the target search service in the first message queue is aggregated according to the time dimension to obtain the aggregated diagnostic data corresponding to each time window of the target search service; The aggregated diagnostic data corresponding to each time window of the target search service is uploaded to the diagnostic database through the asynchronous reporting path.

6. The method of claim 5, wherein, Before the diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, the method further includes: Obtain the current resource consumption corresponding to the target search service; If the current resource consumption corresponding to the target search service is greater than or equal to a resource threshold, the diagnostic data set of the target search service is compressed to obtain diagnostic compressed data of the target search service; The diagnostic data set of the target search service is written into the first message queue through the asynchronous reporting path, including: The diagnostic compressed data of the target search service is written into the first message queue through the asynchronous reporting path.

7. The method of claim 5, wherein, The target search statement carries a search identifier, the first search result, the second search result and the diagnosis information of the target search service of each search resource are summarized to obtain a diagnosis data set of the target search service, comprising: According to the search identifier, obtain the target routing information of the target search service, wherein the target routing information is obtained by hashing the search identifier; The first search result, the second search result and the diagnosis information of the target search service of each search resource are routed to the target node corresponding to the target routing information for summarizing to obtain the diagnosis data set of the target search service.

8. The method of claim 7, wherein, Before the diagnosis data set of the target search service is written into the first message queue through the asynchronous reporting path, the method further comprises: According to the search identifier, index to the delay reporting data and the target node change data; The delay reporting data and the target node change data are summarized to obtain a delay data set; The diagnosis data set of the target search service and the delay data set are written into the first message queue through the asynchronous reporting path. The diagnosis data set of the target search service and the delay data set are written into the first message queue through the asynchronous reporting path.

9. A diagnostic device for search services, characterized in that, Comprise: The acquisition unit is used for acquiring the target search statement for the target search service when the search service diagnosis is needed; The target search statement carries a synchronous reporting parameter and an asynchronous reporting parameter; The processing unit is used for recalling the first search result of each search resource from a plurality of search resources according to the target search statement, wherein the plurality of search resources belong to the search resources associated with the target search service; The processing unit is also used for sorting the first search result of each search resource in a mixed arrangement manner to obtain the second search result of each search resource; The processing unit is also used for diagnosing the processing condition of each search resource in each stage in the plurality of search resources according to the second search result of each search resource to obtain the diagnosis information of the target search service, wherein the diagnosis information includes the first diagnosis information corresponding to the recall of the target search statement in the recall stage and the second diagnosis information corresponding to the sorting of the first search result in the sorting stage; The acquisition unit is also used for acquiring a synchronous reporting path according to the synchronous reporting parameter, wherein the synchronous reporting parameter is used to control the synchronous reporting switch to open the synchronous reporting path; The processing unit is also used for standardizing the first search result, the second search result and the diagnosis information of the target search service of each search resource in a self-defined key-value pair manner to obtain the first standard search result, the second standard search result of each search resource and the standard diagnosis information of the target search service; The first standard search result, the second standard search result and the standard diagnosis information of the target search service of each search resource are uploaded to the diagnosis database through the synchronous reporting path. The acquisition unit is further configured to acquire an asynchronous reporting path according to the asynchronous reporting parameter, wherein the asynchronous reporting parameter is used to control an asynchronous reporting switch to open the asynchronous reporting path. The processing unit is further configured to upload the first search result, the second search result, and the diagnosis information of the target search service of each search resource to the diagnosis database through the asynchronous reporting path. The processing unit is further configured to perform syntax tree analysis on the first standard search result, the second standard search result, and the standard diagnosis information of the target search service of each search resource, to obtain a first analysis result, a second analysis result, and diagnosis analysis information of the target search service of each search resource. The generation unit is configured to generate a search diagnosis list of the target search service according to the first analysis result, the second analysis result, and the diagnosis analysis information of each search resource according to a list generation template.

10. The apparatus of claim 9, wherein, The processing unit is specifically configured to: perform word segmentation processing on the target search statement to obtain at least two to-be-processed words; perform semantic recognition operations on the at least two to-be-processed words respectively to obtain semantic recognition probabilities corresponding to each to-be-processed word; determine an industry category of the target search statement according to the semantic recognition probabilities corresponding to each to-be-processed word; and obtain the first search result of the target search statement in vertical search in each search resource according to the industry category of the target search statement.

11. The apparatus of claim 9, wherein, The processing unit is specifically configured to: read a resource database to obtain a scoring strategy corresponding to each search resource; score the first search result of each search resource according to the scoring strategy to obtain a recall score of each search result in the first search result of each search resource; sort the first search result of each search resource according to the recall scores of the search results to obtain the second search result of each search resource.

12. The apparatus of claim 9, wherein, The processing unit is specifically configured to: write the first search result, the second search result, and the diagnosis information of the target search service of each search resource into a target request packet to obtain a target request packet; upload the target request packet to the diagnosis database through the synchronous reporting path.

13. The apparatus of claim 9, wherein, The processing unit is specifically configured to: aggregate the first search result, the second search result, and the diagnosis information of the target search service of each search resource to obtain a diagnosis data set of the target search service; write the diagnosis data set of the target search service into a first message queue through the asynchronous reporting path; perform aggregation processing on the diagnosis data set of the target search service in the first message queue according to a time dimension to obtain aggregation diagnosis data corresponding to each time window of the target search service; upload the aggregation diagnosis data corresponding to each time window of the target search service to the diagnosis database through the asynchronous reporting path.

14. The apparatus of claim 13, wherein, The acquisition unit is further configured to acquire a current resource consumption of the target search service. The processing unit is further configured to perform data compression on the diagnostic data set of the target search service to obtain diagnostic compressed data of the target search service, if the current resource consumption of the target search service is greater than or equal to the resource threshold. The processing unit is specifically configured to write the diagnostic compressed data of the target search service into a first message queue through the asynchronous reporting path.

15. The apparatus of claim 13, wherein, The target search statement carries a search identifier, and the processing unit is specifically configured to: obtain target routing information of the target search service according to the search identifier, wherein the target routing information is obtained by hashing the search identifier; and route the first search result of each search resource, the second search result, and the diagnostic information of the target search service to a target node corresponding to the target routing information for aggregation to obtain a diagnostic data set of the target search service.

16. The apparatus of claim 15, wherein, The processing unit is further configured to index the delay reporting data and the target node change data according to the search identifier. The processing unit is further configured to aggregate the delay reporting data and the target node change data to obtain a delay data set. The processing unit is specifically configured to write the diagnostic data set of the target search service and the delay data set into a first message queue through the asynchronous reporting path.

17. A computer device, comprising: It comprises: a memory, a transceiver, a processor, and a bus system; The memory is configured to store programs; The processor is configured to execute the programs in the memory to implement the method of any one of claims 1 to 8; The bus system is configured to connect the memory and the processor to enable communication between the memory and the processor.

18. A computer readable storage medium comprising instructions which, when executed on a computer, cause the computer to carry out the method of any one of claims 1 to 8.

19. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instructions implement the steps of the method of any one of claims 1 to 8 when executed by a processor. The computer program / instructions implement the steps of the method of any one of claims 1 to 8 when executed by a processor.

Images